A0439
Title: Tyranny-of-the-minority regression adjustment in randomized experiments
Authors: Hanzhong Liu - Tsinghua University (China) [presenting]
Abstract: Regression adjustment is widely used in the analysis of randomized experiments to improve the estimation efficiency of the treatment effect. A weighted regression adjustment method is termed tyranny-of-the-minority (ToM), wherein units in the minority group are given greater weights. It is demonstrated that ToM regression adjustment is more robust than the previous study's regression adjustment with treatment-covariate interactions, even though these two regression adjustment methods are asymptotically equivalent in completely randomized experiments. Moreover, ToM regression adjustment can be easily extended to stratified randomized experiments and completely randomized survey experiments. The design-based properties of the ToM regression-adjusted average treatment effect estimator are obtained under such designs. In particular, it is shown that the ToM regression-adjusted estimator improves the asymptotic estimation efficiency compared to the unadjusted estimator, even when the regression model is misspecified, and is optimal in the class of linearly adjusted estimators. The asymptotic properties of various heteroscedasticity-robust standard errors are also studied, and recommendations for practitioners are provided. Simulation studies and real data analysis demonstrate ToM regression adjustment's superiority over existing methods.
